1. Technical Field
Present invention embodiments relate to determining relationships between database objects, and more specifically, to determining pivot type (e.g., reverse pivot, pivot, etc.) relationships between database objects.
2. Discussion of the Related Art
Pivoting or reverse pivoting of database tables is a technology in data mining and data warehousing intended to increase efficiency of data warehousing, data mining, and other data analysis. Data is pivoted or reverse pivoted between source and target data sets. For example, data of a target database table column (e.g., indicating a particular quarter of a year, etc.) may serve as columns for the source database table (e.g., for data pertaining to that quarter).
Discovery of relationships between source and target data sets is an initial stage of information technology projects. This discovery may be automated, where relationships between source and target data are analyzed to determine mappings designating the manner in which the target data is generated from the source data. The mappings include binding conditions and transformations. The binding conditions pertain to the conditions (expressed as SQL) that relate portions of the source database table to corresponding portions of the target database table. The transformations include expressions (SQL) for each target database table column that comprise source database table columns, constants, and valid SQL expressions. The transformation expressions define the manner in which target database table column data is generated from source database table columns related by a binding condition. However, the binding conditions and transformations are not sufficient to analyze reverse pivot relationships since the target data cannot be re-constructed with a single mapping.